Coverage Report

Created: 2026-09-14 07:15

next uncovered line (L), next uncovered region (R), next uncovered branch (B)
/src/serenity/Userland/Libraries/LibAudio/WavLoader.cpp
Line
Count
Source
1
/*
2
 * Copyright (c) 2018-2020, Andreas Kling <kling@serenityos.org>
3
 * Copyright (c) 2021-2023, kleines Filmröllchen <filmroellchen@serenityos.org>
4
 *
5
 * SPDX-License-Identifier: BSD-2-Clause
6
 */
7
8
#include "WavLoader.h"
9
#include "LoaderError.h"
10
#include "WavTypes.h"
11
#include <AK/Debug.h>
12
#include <AK/Endian.h>
13
#include <AK/FixedArray.h>
14
#include <AK/MemoryStream.h>
15
#include <AK/NonnullOwnPtr.h>
16
#include <AK/NumericLimits.h>
17
#include <AK/Try.h>
18
19
namespace Audio {
20
21
WavLoaderPlugin::WavLoaderPlugin(NonnullOwnPtr<SeekableStream> stream)
22
0
    : LoaderPlugin(move(stream))
23
0
{
24
0
}
25
26
bool WavLoaderPlugin::sniff(SeekableStream& stream)
27
0
{
28
0
    auto riff = stream.read_value<RIFF::ChunkID>();
29
0
    if (riff.is_error())
30
0
        return false;
31
0
    if (riff.value() != RIFF::riff_magic)
32
0
        return false;
33
34
0
    auto size = stream.read_value<LittleEndian<u32>>();
35
0
    if (size.is_error())
36
0
        return false;
37
38
0
    auto wave = stream.read_value<RIFF::ChunkID>();
39
0
    return !wave.is_error() && wave.value() == Wav::wave_subformat_id;
40
0
}
41
42
ErrorOr<NonnullOwnPtr<LoaderPlugin>, LoaderError> WavLoaderPlugin::create(NonnullOwnPtr<SeekableStream> stream)
43
0
{
44
0
    auto loader = make<WavLoaderPlugin>(move(stream));
45
0
    TRY(loader->parse_header());
46
0
    return loader;
47
0
}
48
49
template<typename SampleReader>
50
MaybeLoaderError WavLoaderPlugin::read_samples_from_stream(Stream& stream, SampleReader read_sample, FixedArray<Sample>& samples) const
51
0
{
52
0
    switch (m_num_channels) {
53
0
    case 1:
54
0
        for (auto& sample : samples)
55
0
            sample = Sample(TRY(read_sample(stream)));
56
0
        break;
57
0
    case 2:
58
0
        for (auto& sample : samples) {
59
0
            auto left_channel_sample = TRY(read_sample(stream));
60
0
            auto right_channel_sample = TRY(read_sample(stream));
61
0
            sample = Sample(left_channel_sample, right_channel_sample);
62
0
        }
63
0
        break;
64
0
    default:
65
0
        VERIFY_NOT_REACHED();
66
0
    }
67
0
    return {};
68
0
}
69
70
// There's no i24 type + we need to do the endianness conversion manually anyways.
71
static ErrorOr<double> read_sample_int24(Stream& stream)
72
0
{
73
0
    i32 sample1 = TRY(stream.read_value<u8>());
74
0
    i32 sample2 = TRY(stream.read_value<u8>());
75
0
    i32 sample3 = TRY(stream.read_value<u8>());
76
77
0
    i32 value = 0;
78
0
    value = sample1;
79
0
    value |= sample2 << 8;
80
0
    value |= sample3 << 16;
81
    // Sign extend the value, as it can currently not have the correct sign.
82
0
    value = (value << 8) >> 8;
83
    // Range of value is now -2^23 to 2^23-1 and we can rescale normally.
84
0
    return static_cast<double>(value) / static_cast<double>((1 << 23) - 1);
85
0
}
86
87
template<typename T>
88
static ErrorOr<double> read_sample(Stream& stream)
89
0
{
90
0
    T sample { 0 };
91
0
    TRY(stream.read_until_filled(Bytes { &sample, sizeof(T) }));
92
    // Remap integer samples to normalized floating-point range of -1 to 1.
93
0
    if constexpr (IsIntegral<T>) {
94
0
        if constexpr (NumericLimits<T>::is_signed()) {
95
            // Signed integer samples are centered around zero, so this division is enough.
96
0
            return static_cast<double>(AK::convert_between_host_and_little_endian(sample)) / static_cast<double>(NumericLimits<T>::max());
97
0
        } else {
98
            // Unsigned integer samples, on the other hand, need to be shifted to center them around zero.
99
            // The first division therefore remaps to the range 0 to 2.
100
0
            return static_cast<double>(AK::convert_between_host_and_little_endian(sample)) / (static_cast<double>(NumericLimits<T>::max()) / 2.0) - 1.0;
101
0
        }
102
0
    } else {
103
0
        return static_cast<double>(AK::convert_between_host_and_little_endian(sample));
104
0
    }
105
0
}
Unexecuted instantiation: WavLoader.cpp:AK::ErrorOr<double, AK::Error> Audio::read_sample<unsigned char>(AK::Stream&)
Unexecuted instantiation: WavLoader.cpp:AK::ErrorOr<double, AK::Error> Audio::read_sample<short>(AK::Stream&)
Unexecuted instantiation: WavLoader.cpp:AK::ErrorOr<double, AK::Error> Audio::read_sample<float>(AK::Stream&)
Unexecuted instantiation: WavLoader.cpp:AK::ErrorOr<double, AK::Error> Audio::read_sample<double>(AK::Stream&)
106
107
LoaderSamples WavLoaderPlugin::samples_from_pcm_data(ReadonlyBytes data, size_t samples_to_read) const
108
0
{
109
0
    FixedArray<Sample> samples = TRY(FixedArray<Sample>::create(samples_to_read));
110
0
    FixedMemoryStream stream { data };
111
112
0
    switch (m_sample_format) {
113
0
    case PcmSampleFormat::Uint8:
114
0
        TRY(read_samples_from_stream(stream, read_sample<u8>, samples));
115
0
        break;
116
0
    case PcmSampleFormat::Int16:
117
0
        TRY(read_samples_from_stream(stream, read_sample<i16>, samples));
118
0
        break;
119
0
    case PcmSampleFormat::Int24:
120
0
        TRY(read_samples_from_stream(stream, read_sample_int24, samples));
121
0
        break;
122
0
    case PcmSampleFormat::Float32:
123
0
        TRY(read_samples_from_stream(stream, read_sample<float>, samples));
124
0
        break;
125
0
    case PcmSampleFormat::Float64:
126
0
        TRY(read_samples_from_stream(stream, read_sample<double>, samples));
127
0
        break;
128
0
    default:
129
0
        VERIFY_NOT_REACHED();
130
0
    }
131
132
0
    return samples;
133
0
}
134
135
ErrorOr<Vector<FixedArray<Sample>>, LoaderError> WavLoaderPlugin::load_chunks(size_t samples_to_read_from_input)
136
0
{
137
0
    auto remaining_samples = m_total_samples - m_loaded_samples;
138
0
    if (remaining_samples <= 0)
139
0
        return Vector<FixedArray<Sample>> {};
140
141
    // One "sample" contains data from all channels.
142
    // In the Wave spec, this is also called a block.
143
0
    size_t bytes_per_sample
144
0
        = m_num_channels * pcm_bits_per_sample(m_sample_format) / 8;
145
146
0
    auto samples_to_read = min(samples_to_read_from_input, remaining_samples);
147
0
    auto bytes_to_read = samples_to_read * bytes_per_sample;
148
149
0
    dbgln_if(AWAVLOADER_DEBUG, "Read {} bytes WAV with num_channels {} sample rate {}, "
150
0
                               "bits per sample {}, sample format {}",
151
0
        bytes_to_read, m_num_channels, m_sample_rate,
152
0
        pcm_bits_per_sample(m_sample_format), sample_format_name(m_sample_format));
153
154
0
    auto sample_data = TRY(ByteBuffer::create_zeroed(bytes_to_read));
155
0
    TRY(m_stream->read_until_filled(sample_data.bytes()));
156
157
    // m_loaded_samples should contain the amount of actually loaded samples
158
0
    m_loaded_samples += samples_to_read;
159
0
    Vector<FixedArray<Sample>> samples;
160
0
    TRY(samples.try_append(TRY(samples_from_pcm_data(sample_data.bytes(), samples_to_read))));
161
0
    return samples;
162
0
}
163
164
MaybeLoaderError WavLoaderPlugin::seek(int sample_index)
165
0
{
166
0
    dbgln_if(AWAVLOADER_DEBUG, "seek sample_index {}", sample_index);
167
0
    if (sample_index < 0 || sample_index >= static_cast<int>(m_total_samples))
168
0
        return LoaderError { LoaderError::Category::Internal, m_loaded_samples, "Seek outside the sample range"_fly_string };
169
170
0
    size_t sample_offset = m_byte_offset_of_data_samples + static_cast<size_t>(sample_index * m_num_channels * (pcm_bits_per_sample(m_sample_format) / 8));
171
172
0
    TRY(m_stream->seek(sample_offset, SeekMode::SetPosition));
173
174
0
    m_loaded_samples = sample_index;
175
0
    return {};
176
0
}
177
178
// Specification reference: http://www-mmsp.ece.mcgill.ca/Documents/AudioFormats/WAVE/WAVE.html
179
MaybeLoaderError WavLoaderPlugin::parse_header()
180
0
{
181
0
#define CHECK(check, category, msg)                                                                                                     \
182
0
    do {                                                                                                                                \
183
0
        if (!(check)) {                                                                                                                 \
184
0
            return LoaderError { category, static_cast<size_t>(TRY(m_stream->tell())), TRY(String::formatted("WAV header: {}", msg)) }; \
185
0
        }                                                                                                                               \
186
0
    } while (0)
187
188
0
    auto file_header = TRY(m_stream->read_value<RIFF::FileHeader>());
189
0
    CHECK(file_header.magic() == RIFF::riff_magic, LoaderError::Category::Format, "RIFF header magic invalid");
190
0
    CHECK(file_header.subformat == Wav::wave_subformat_id, LoaderError::Category::Format, "WAVE subformat id invalid");
191
192
0
    auto format_chunk = TRY(m_stream->read_value<RIFF::OwnedChunk>());
193
0
    CHECK(format_chunk.id().as_ascii_string() == Wav::format_chunk_id, LoaderError::Category::Format, "FMT chunk id invalid");
194
195
0
    auto format_stream = format_chunk.data_stream();
196
0
    u16 audio_format = TRY(format_stream.read_value<LittleEndian<u16>>());
197
0
    CHECK(audio_format == to_underlying(Wav::WaveFormat::Pcm) || audio_format == to_underlying(Wav::WaveFormat::IEEEFloat) || audio_format == to_underlying(Wav::WaveFormat::Extensible),
198
0
        LoaderError::Category::Unimplemented, "Audio format not supported");
199
200
0
    m_num_channels = TRY(format_stream.read_value<LittleEndian<u16>>());
201
0
    CHECK(m_num_channels == 1 || m_num_channels == 2, LoaderError::Category::Unimplemented, "Channel count");
202
203
0
    m_sample_rate = TRY(format_stream.read_value<LittleEndian<u32>>());
204
    // Data rate; can be ignored.
205
0
    TRY(format_stream.read_value<LittleEndian<u32>>());
206
0
    u16 block_size_bytes = TRY(format_stream.read_value<LittleEndian<u16>>());
207
208
0
    u16 bits_per_sample = TRY(format_stream.read_value<LittleEndian<u16>>());
209
210
0
    if (audio_format == to_underlying(Wav::WaveFormat::Extensible)) {
211
0
        CHECK(format_chunk.size() == 40, LoaderError::Category::Format, "Extensible fmt size is not 40 bytes");
212
213
        // Discard everything until the GUID.
214
        // We've already read 16 bytes from the stream. The GUID starts in another 8 bytes.
215
0
        TRY(format_stream.read_value<LittleEndian<u64>>());
216
217
        // Get the underlying audio format from the first two bytes of GUID
218
0
        u16 guid_subformat = TRY(format_stream.read_value<LittleEndian<u16>>());
219
0
        CHECK(guid_subformat == to_underlying(Wav::WaveFormat::Pcm) || guid_subformat == to_underlying(Wav::WaveFormat::IEEEFloat), LoaderError::Category::Unimplemented, "GUID SubFormat not supported");
220
221
0
        audio_format = guid_subformat;
222
0
    }
223
224
0
    if (audio_format == to_underlying(Wav::WaveFormat::Pcm)) {
225
0
        CHECK(bits_per_sample == 8 || bits_per_sample == 16 || bits_per_sample == 24, LoaderError::Category::Unimplemented, "PCM bits per sample not supported");
226
227
        // We only support 8-24 bit audio right now because other formats are uncommon
228
0
        if (bits_per_sample == 8) {
229
0
            m_sample_format = PcmSampleFormat::Uint8;
230
0
        } else if (bits_per_sample == 16) {
231
0
            m_sample_format = PcmSampleFormat::Int16;
232
0
        } else if (bits_per_sample == 24) {
233
0
            m_sample_format = PcmSampleFormat::Int24;
234
0
        }
235
0
    } else if (audio_format == to_underlying(Wav::WaveFormat::IEEEFloat)) {
236
0
        CHECK(bits_per_sample == 32 || bits_per_sample == 64, LoaderError::Category::Unimplemented, "Float bits per sample not supported");
237
238
        // Again, only the common 32 and 64 bit
239
0
        if (bits_per_sample == 32) {
240
0
            m_sample_format = PcmSampleFormat::Float32;
241
0
        } else if (bits_per_sample == 64) {
242
0
            m_sample_format = PcmSampleFormat::Float64;
243
0
        }
244
0
    }
245
246
0
    CHECK(block_size_bytes == (m_num_channels * (bits_per_sample / 8)), LoaderError::Category::Format, "Block size invalid");
247
248
0
    dbgln_if(AWAVLOADER_DEBUG, "WAV format {} at {} bit, {} channels, rate {}Hz ",
249
0
        sample_format_name(m_sample_format), pcm_bits_per_sample(m_sample_format), m_num_channels, m_sample_rate);
250
251
    // Read all chunks before DATA.
252
0
    bool found_data = false;
253
0
    while (!found_data) {
254
0
        auto chunk_header = TRY(m_stream->read_value<RIFF::ChunkID>());
255
0
        if (chunk_header == Wav::data_chunk_id) {
256
0
            found_data = true;
257
0
        } else {
258
0
            TRY(m_stream->seek(-RIFF::chunk_id_size, SeekMode::FromCurrentPosition));
259
0
            auto chunk = TRY(m_stream->read_value<RIFF::OwnedChunk>());
260
0
            if (chunk.id() == RIFF::list_chunk_id) {
261
0
                auto maybe_list = chunk.data_stream().read_value<RIFF::OwnedList>();
262
0
                if (maybe_list.is_error()) {
263
0
                    dbgln("WAV Warning: LIST chunk invalid, error: {}", maybe_list.release_error());
264
0
                    continue;
265
0
                }
266
267
0
                auto list = maybe_list.release_value();
268
0
                if (list.type == Wav::info_chunk_id) {
269
0
                    auto maybe_error = load_wav_info_block(move(list.chunks));
270
0
                    if (maybe_error.is_error())
271
0
                        dbgln("WAV Warning: INFO chunk invalid, error: {}", maybe_error.release_error());
272
273
0
                } else {
274
0
                    dbgln("Unhandled WAV list of type {} with {} subchunks", list.type.as_ascii_string(), list.chunks.size());
275
0
                }
276
0
            } else {
277
0
                dbgln_if(AWAVLOADER_DEBUG, "Unhandled WAV chunk of type {}, size {} bytes", chunk.id().as_ascii_string(), chunk.size());
278
0
            }
279
0
        }
280
0
    }
281
282
0
    u32 data_size = TRY(m_stream->read_value<LittleEndian<u32>>());
283
0
    CHECK(found_data, LoaderError::Category::Format, "Found no data chunk");
284
285
0
    m_total_samples = data_size / block_size_bytes;
286
287
0
    dbgln_if(AWAVLOADER_DEBUG, "WAV data size {}, bytes per sample {}, total samples {}",
288
0
        data_size,
289
0
        block_size_bytes,
290
0
        m_total_samples);
291
292
0
    m_byte_offset_of_data_samples = TRY(m_stream->tell());
293
0
    return {};
294
0
}
295
296
// http://www-mmsp.ece.mcgill.ca/Documents/AudioFormats/WAVE/Docs/riffmci.pdf page 23 (LIST type)
297
// We only recognize the relevant official metadata types; types added in later errata of RIFF are not relevant for audio.
298
MaybeLoaderError WavLoaderPlugin::load_wav_info_block(Vector<RIFF::OwnedChunk> info_chunks)
299
0
{
300
0
    for (auto const& chunk : info_chunks) {
301
0
        auto chunk_id = chunk.id();
302
0
        auto metadata_name = chunk_id.as_ascii_string();
303
        // Chunk contents are zero-terminated strings "ZSTR", so we just drop the null terminator.
304
0
        StringView metadata_text { chunk.data().trim(chunk.size() - 1) };
305
        // Note that we assume chunks to be unique, since that seems to almost always be the case.
306
        // Worst case we just drop some metadata.
307
0
        if (metadata_name == "IART"sv) {
308
            // Artists are combined together with semicolons, at least when you edit them in Windows File Explorer.
309
0
            auto artists = metadata_text.split_view(";"sv);
310
0
            for (auto artist : artists)
311
0
                TRY(m_metadata.add_person(Person::Role::Artist, TRY(String::from_utf8(artist))));
312
0
        } else if (metadata_name == "ICMT"sv) {
313
0
            m_metadata.comment = TRY(String::from_utf8(metadata_text));
314
0
        } else if (metadata_name == "ICOP"sv) {
315
0
            m_metadata.copyright = TRY(String::from_utf8(metadata_text));
316
0
        } else if (metadata_name == "ICRD"sv) {
317
0
            m_metadata.unparsed_time = TRY(String::from_utf8(metadata_text));
318
0
        } else if (metadata_name == "IENG"sv) {
319
0
            TRY(m_metadata.add_person(Person::Role::Engineer, TRY(String::from_utf8(metadata_text))));
320
0
        } else if (metadata_name == "IGNR"sv) {
321
0
            m_metadata.genre = TRY(String::from_utf8(metadata_text));
322
0
        } else if (metadata_name == "INAM"sv) {
323
0
            m_metadata.title = TRY(String::from_utf8(metadata_text));
324
0
        } else if (metadata_name == "IPRD"sv) {
325
0
            m_metadata.album = TRY(String::from_utf8(metadata_text));
326
0
        } else if (metadata_name == "ISFT"sv) {
327
0
            m_metadata.encoder = TRY(String::from_utf8(metadata_text));
328
0
        } else if (metadata_name == "ISRC"sv) {
329
0
            TRY(m_metadata.add_person(Person::Role::Publisher, TRY(String::from_utf8(metadata_text))));
330
0
        } else {
331
0
            TRY(m_metadata.add_miscellaneous(TRY(String::from_utf8(metadata_name)), TRY(String::from_utf8(metadata_text))));
332
0
        }
333
0
    }
334
0
    return {};
335
0
}
336
337
}